THECYCLIZATION OF NITRILES
May 20, 1956
complex mixtures, fractions 2, 3 and 4 ( n ~1.5930, , 1.5930, 1.5927) were combined and fractionated under very high reflux a t 70 mm. to yield the constant volume (4 ml.) fractions shown in Table 11.
TABLE I1 FractiOn
ltUD
1 2
1.5823 1.5885 1.5896 1.5908 1.5916 1.5921 1.5929
3 4 5 6 7
Picrate, mp, C.
Fraction
Not prepd. 8 9 240 dec. 10 245 dec. 11 243 dec. 237-239 dec. 12 241 dec. 13 245 dec. Residue
Picrate, m.p., n25D
1.5932 1.5937 1.5940 1.5948 1.5950 1.5950
OC.
240 dec. 200-207 241-243 220-222 252-253 210-215
This table is presented to indicate the wide range of melting points of the picrates obtained from these cuts. Of the known quinoline and isoquinoline homologs none has been reported with a picrate melting as high as 245" SO a number of the picrates shown must be those of bases not previously isolated from petroleum or converted to picrate after synthesis. The picrates from cuts 2 and 8 showed no depression in mixed melting point so these and intervening cuts were combined and studied. The main base was not identical with any known base and the amount liberated from 18 g. of purified picrate did not permit the extensive degradation that would be needed before deciding what base should be synthesized and compared with the shale-oil base. Similar fractionation of neutralization cuts 9, 10 and 11 yielded another series of fractions all of which could easily be converted to solid picrates but only the well known, 2,3,8-trimethylquinoline could be identified. A number of other apparently pure picrates were obtained but none were identified. The identity of 2,3,8-trimethylquinoline was established by analysis and by comparison of its properties with those of the compound reported by Poth, et aZ.ls (calculated or previously reported properties in parentheses): m.p. 51-52"
THE
(55-56'); equivalent weight, non-aqueous titration, 171.5 (calcd. 171.23); % N, micro-Dumas, 8.23 (calcd., 8.19); m.p. Dicrate, 242-243' (243"); sulfate, 275'dec. (275'dec.). MixLd melting points were not depressed below the lower values. In the search for the CIGH~SN. chloroform layer base 530 ml. of bases isolated from the chloroform phase were distilled from a Claisen flask, yield 480 ml. of bases, 50 ml. of tar. Careful fractionation through the spinning band column a t 75 mm. pressure yielded 29 equal volume cuts and a 20-ml. tarry residue. The boiling range was 265 to 308' and the index of refraction range to 1.5200 a t cut 18 and a final value of 1.5360. The Cl6 base from getroleurn has b.p. 278"; %*OD 1.5129; and picrate m.p. 151 Fractions 5 and 6 were selected as the ones that should contain the highest concentration of the base, if present, and attempts were made to isolate the solid picrate, hydrochloride, or acid sulfate, by the various methods employed by Bailey and s t ~ d e n t s ~ ~ Jin~ Jtheir 0 prolonged study of this base. Many of their fractions were less highly fractionated than the shale-oil base a t this stage but no solid derivative could be isolated. Similar tests on California petroleum base fractions yielded the solid derivatives without difficulty. Since a base b.p. 278" might be expected to be present in highest concentration in fractions somewhat lower boiling than fractions 5 and 6 , another worker, Tom Cheavens, studied cuts 2 to 4 and again was unable to isolate a solid derivative, so it must be concluded that this interesting base, if present in shale-oil bases, is present in very low concentration or is mixed with types of bas,es from which it cannot be obtained in the form of solid derivatives without extensive additional fractionation probably by different methods.
.
Acknowledgment.-The authors wish to thank General Aniline and Film Corporation for supporting this work through fellowships from 1950-1953, and securing the generous supply of material for study. (19) W. C. Thompson and J. R. Bailey, i b i d . , 63, 1002 (1931). (20) B. P. Armendt and J. R . Bailey, ibid., 66, 4143 (1933).
AUSTIN. TEXAS
(18) E. J. Poth, et al., THIS JOURNAL, 62, 1239 (1930).
[CONTRIBUTION FROM
2153
DEPARTMENT OF CHEMISTRY, DUKEUNIVERSITY]
The Cyclization of Nitriles : A New Route to Some Phenanthrylarnine~l-~ BY C. K. BRADSHER, E. D. LITTLE AND DOROTHY J. BEAVERS RECEIVED NOVEVBER 19, 1956 I n the presence of concentrated sulfuric acid, (2-biphenyly1)-acetonitrile (VI, R = H). as well as certain a-alkyl and LYaroyl derivatives, undergo cyclization readily t o yield the isomeric 9-phenanthrylamines (VII). A compound previously described as 9-(p-methoxyphenyl)-lO-phenanthramide(IIIA) has been shown to be lO-(p-methoxybenzoyl)-9-phenanthrylamine (111).
In earlier work4 it was shown that a-(p-methoxybenzoyl) -a-(o-biphenylyl)-acetonitrile (I) in the presence of concentrated sulfuric acid underwent conversion to an isomeric product. This product from analogy to the behavior of the phenyl analog of I (I, R = H) and from a consideration of certain features of the absorption spectrum, appeared to be IIIA. As part of a program for the synthesis of 9-phen(1) This work has been the subject of a preliminary communication, THISJOURNAL, 76, 948 (1954). ( 2 ) This investigation was supported by a research grant (C-1743) from the National Cancer Institute of the National Institutes of Health, Public Health Service. (3) Abstracted in part from theses submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (D.J.B.) and Master of Arts (E.D.L.). (4) C. K. Bradsher and R . S. Kittila, THISJOURNAL,71, 277 (1950).
ylphenanthrene derivatives5 it seemed desirable to prepare a sample of the isomeric 9-(o-methoxyphenyl)-10-phenanthramide(IVA). The acylation of (2-biphenylyl)-acetonitrile with methyl o-methoxybenzoate afforded I1 in 46% yield. When I1 was dissolved in concentrated sulfuric acid a t 0" and allowed to remain a t this temperature for three hours, a product having the composition expected for IVA was obtained in a yield of 85%. A sample of the new product was refluxed with hydrobromic and acetic acids with the idea that ether cleavage and hydrolysis might yield the lactone of 9-(o-hydroxyphenyl)-lO-phenanthroic (5) Triphenylethylene derivatives, some of which have been found active as tumor necrotizing agents (e.g., G . M. Badger, L. A. Elson, A. Haddow, C. L. Hewett and M. Robinson, Proc. R o y . SOC.(London), B180, 255 (1941)), may be regarded as "open" models of 9-phenylphenanthrene derivatives.
2154
C. K. BRADSHER, E. D. LITTLEAND D. J. BEAVERS
VOl. 78
The results obtained in the cyclization of these to yield 9-alkyl-10-phenanthrylamines (1711) are summarized in Table I. TABLE I (1.11)
PHENANTHKYLANISES
B Y CYCLIZATION OF S I I ' K I L E S
Yield,a
%
Amine
72-85 61 83 74 62
138-138.5' 148.5-150 117.5-119.5 101-102.5 70.5-72
K \/'
111, X = P-OCH, IT-,X o-OCH~
\.
acid. Actually a mixture was obtained from which the only product identified was salicylic acid. Clearly the cleavage of salicylic acid from a molecule having the structure IVA would be very unlikely. Xn alternate possibility was that the nitrile rather than the carbonyl group of I1 had undergone an acid-catalyzed reaction with the adjacent phenyl nucleus yielding 9-amino-10-(o-methoxybenzoyl)phenanthrene (IV). I n support of this formulation for IV, it was observed that an acetic acid solution could be diazotized to give a positive diazonium coupling test with /3-naphthol. These observations led to a reconsideration of the structure IILI assigned earlier4 to the cyclization product obtained from I . It had already been demonstrated that this product reacted with nitrous acid to yield an unknown compound having the composition expected for C22H1402. It has now been found that the product from I gives a characteristic diazonium coupling test. In the light of this, the product from I must be represented as I11 rather than IIIX. The ability of diazotized derivatives of 2-aminobenzophenone to undergo cyclization6 to yield fluorene derivatives suggests that the previously unidentified product (C22H1402) obtained from I11 by the action of nitrous acid is 1,2,3,4-dibenzo-6-methoxyfluorenone (V) . X logical extension of the new cyclization reaction would be its application to simple a-(2-biphenyly1)-nitriles (VI). Although the first member o f this series (R = H) has been known for several years7 there is no record of a cyclization being xttempted. When 2-biphenylylacetonitrile (VII, R = H) was dissolved in concentrated sulfuric acid and allowed to stand a t 0" for three hours, yields of 72-8570 of 9-phenanthrylamine (1711, R = H) were obtained. Some of the homologs of 2-biphenylylacetonitrile (VI, R = H) have been prepared from it by alkylation using sodiuin amide and an alkyl halide.8 C ( h i e h e ; i n d F. Y l l r n a n r i , / $ c , r , , 27, 3485 i l 8 0 1 ) ; I:. [~Ilrn,in R l e i r r , i b i d . , 35, 4275 (lWl'2). 7 ) J . v o n Braiio and G . l l a n z , A n i i . , 468, L'i8 (I!J291. X j C . K . Bradsher and 1". J Jxrkson, J r , T r r r s J o i i K % % I . , 73, 32:iT) ( l ! ~ > l )~ ;b i ~ 7f 6, , 731 ( l ! l > 4 , . 'It]
; i i i < l 13.
I€ CHa C2Hj CBH, C4H9
l l . p . , b OC. hydrochloride
Benzamide
278 (dec.)d 199-201e 2 7 3 . 5 2 7 7 251-252 268-270 241-243 250.5-252 220-221.5 275-277 . . . . . .f
Yields are of products melting within 3.5" of the analytical sample. Reported melting points are for analytical samples. C Reported 137.5-138' by L. F. Fieser, R . P . Jacobson and C. C. Price, THISJOURNAL, 5 8 , 2163 (1936). Reported CQ. 275" dec. by J. Schmidt and M . Strobel, R c r . , 34, 1461 (1901). EReported 199" by W. E. Bachmann and C. H . Boatner, THISJOURNAL, 58, 2097 (1936). 1 Converted t o the dibenzoyl derivative, m.p. 171-172'.
Although no effort was made to find the optimum conditions for each cyclization, the yields were quite good for the compounds listed. If a-(2biphenylyl) -p-phenylpropionitrile (VI, R = CeHjCHz), was subjected to the usual cyclization conditions sulfonation appeared to be the chief reaction. A similar observation was made with 2-phenyl-4inethoxybenzyl ~ y a n i d e . ~ On the other hand (4,4'dibromo-2-biphenylyl)-acetonitrile10 (VIII) in which the aromatic rings are strongly deactivated failed to cyclize, and gave instead the corresponding amide IX.'O Br
fi L'\CH?
fix I
\/
Br VIII, 9 IX, X
'v'
= =
cs
s
SI
COSHz
Dialkylation of biphenylylacetonitrile (VI), K = H) with butyl bromide afforded a nitrile X which on cyclization in the usual way gave the known'' 10,lO-
dibuty1-9-keto-9,lO-dihydrophenanthrene(XI). The new cyclization reaction is clearly an intra(!))
C. K.
13r;rdsher
and
\T. .I.Jackson, J r . , ;hid., 7 4 , 4880
fl9>2J.
c.K .
13ra
